CN113592914B - Infrared weak and small flying target self-adaptive detection tracking method and device - Google Patents
Infrared weak and small flying target self-adaptive detection tracking method and device Download PDFInfo
- Publication number
- CN113592914B CN113592914B CN202111142157.4A CN202111142157A CN113592914B CN 113592914 B CN113592914 B CN 113592914B CN 202111142157 A CN202111142157 A CN 202111142157A CN 113592914 B CN113592914 B CN 113592914B
- Authority
- CN
- China
- Prior art keywords
- window
- size
- template
- local
- target
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/20—Analysis of motion
- G06T7/246—Analysis of motion using feature-based methods, e.g. the tracking of corners or segments
- G06T7/251—Analysis of motion using feature-based methods, e.g. the tracking of corners or segments involving models
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10048—Infrared image
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention discloses an infrared small and weak flying target self-adaptive detection tracking method and device, wherein the method comprises the following steps: the image frame of the original infrared image is subjected to a window size adjustment step, and a first variance map and a second variance map are obtained by calculating through the image frame by using a first template and a second template; wherein, the window size adjusting step: calculating the global standard deviation of the image frame; traversing the image frame by using a first template, calculating to obtain a first local standard deviation, increasing the size of a first window until the global standard deviation is larger than the first local standard deviation, and determining the size of the first window; and traversing the image frame by using a second template, calculating to obtain a second local standard deviation, increasing the size of the second window until the second local standard deviation is larger than the first local standard deviation, and determining the size of the second window. By adopting the technical scheme, the size of the filtering window is changed in a self-adaptive manner, and based on the changed filtering window, the weak and small targets are stably captured constantly, so that the reliability of detecting and tracking the weak and small targets is improved.
Description
Technical Field
The invention relates to the technical field of target detection and tracking, in particular to an infrared weak and small flying target self-adaptive detection and tracking method and device.
Background
Because the small and weak flying targets are very small in the infrared image, the detection and tracking of the small and weak flying targets are also interfered by noise, clutter or cloud layers, so the signal-to-noise ratio of the small and weak flying targets is usually very low and often submerged in the background.
In the prior art, although the scheme for detecting and tracking the weak and small targets overcomes the difficulty of large computation amount to a certain extent, the problem still exists that the size of an applied filtering window is fixed, and the size variability of the flying targets on the image is ignored. When the target is far away from the detector, the target presents point target characteristics, and when the target is close to the detector, the target presents surface source characteristics. Based on the weak and small characteristics of the target, spatial information such as shape, size, dimension, texture and the like is unavailable, and along with the change of the imaging size of the target, when the imaging size of the target is larger than a filtering window or far smaller than the filtering window, the target detection and tracking are easy to fail.
Disclosure of Invention
The purpose of the invention is as follows: the invention provides an infrared small and weak flying target self-adaptive detection tracking method and device, aiming at realizing the purpose of stably capturing small and weak targets at any moment and improving the reliability of small and weak target detection tracking according to the change of the imaging size of the small and weak targets and the corresponding self-adaptive change of a filter window.
The technical scheme is as follows: the invention provides an infrared small and weak flying target self-adaptive detection tracking method, which comprises the following steps:
acquiring an original infrared image including a target to be detected;
the image frame of the original infrared image is subjected to window size adjustment, a first template and a second template are suitable for traversing the image frame for calculation, and the variance of the gray value in the range of the first template is calculated to obtain a first variance map; calculating the variance of the gray value in the second template range to obtain a second variance map; the range of the first template is a first window, and the range of the second template is the range of subtracting a second window positioned in the first window from the first window;
wherein, the window size adjusting step: calculating the global standard deviation of the image frame; traversing the image frame by using a first template, calculating to obtain a plurality of first local standard deviations, if the global standard deviation is not larger than the first local standard deviation, increasing the size of a first window until the global standard deviation is larger than the first local standard deviation, and determining the size of the first window; traversing the image frame by using a second template, calculating to obtain a plurality of second local standard deviations, if the second local standard deviations are not larger than the first local standard deviations, increasing the size of a second window until the second local standard deviations are larger than the first local standard deviations, and determining the size of the second window;
subtracting the second variogram from the first variogram to obtain a variogram, and selecting a maximum value point as a suspected target point;
and tracking the target to be detected according to the suspected target point.
Specifically, the size of the first window and the size of the second window are adjusted by sequentially applying a window size adjustment step to each image frame of the original infrared image.
Specifically, the variance of the gray value obtained by calculation is used for replacing the gray value of the center point of the template to obtain a variance map.
Specifically, the second template, the first window and the second window share a common center point.
Specifically, the initial sizes of the first window and the second window are N pixel points, and N is more than or equal to 3;
if Sstd_total＞λS1std_localThen the current first window size is determined, otherwise the first window size L = (N + 2N), N =1, 2, 3, … …, Sstd_totalIndicating the global standard deviation, S1std_localDenotes a first local standard deviation, λ denotes a tuning parameter;
if S2std_local＞λS1std_localThen the current second window size is determined, otherwise the second window size L = (N + 2N), N =1, 2, 3, … …, S2std_localThe second local standard deviation is indicated.
Specifically, for each image frame of the original infrared image, the variance distribution of the gray values of the cloud layer edge area, the cloud layer area and the clearance area is calculated, and the adjustment parameters in the step of adjusting the size of the application window are adjusted according to the variance distribution of each image frame.
Specifically, spatial filtering is performed on the image frame of the original infrared image, and the filtered image frame is multiplied by corresponding pixel points of the variance difference map to obtain a background inhibition map;
screening a preset number of maximum value points in the background inhibition map as suspected target points;
and carrying out target segmentation according to the suspected target point to obtain a suspected target.
Specifically, a suspected target is tracked by using a wave gate, and if the suspected target is in the wave gate range, the suspected target is determined to be a target to be detected; and if the suspected target is out of the wave gate range, judging the suspected target to be a new target or a false alarm.
The invention also provides an infrared small and weak flying target self-adaptive detection tracking device, which comprises an acquisition unit, a variance calculation unit, a window adjustment unit, a screening unit and a tracking unit, wherein:
the acquisition unit is used for acquiring an original infrared image comprising a target to be detected;
the variance calculating unit is used for applying a window size adjusting step to the image frames of the original infrared images, calculating by traversing the image frames by using a first template and a second template, and calculating the variance of the gray value in the range of the first template to obtain a first variance map; calculating the variance of the gray value in the second template range to obtain a second variance map; the range of the first template is a first window, and the range of the second template is the range of subtracting a second window positioned in the first window from the first window;
the window adjusting unit is used for adjusting the window size: calculating the global standard deviation of the image frame; traversing the image frame by using a first template, calculating to obtain a plurality of first local standard deviations, if the global standard deviation is not larger than the first local standard deviation, increasing the size of a first window until the global standard deviation is larger than the first local standard deviation, and determining the size of the first window; traversing the image frame by using a second template, calculating to obtain a plurality of second local standard deviations, if the second local standard deviations are not larger than the first local standard deviations, increasing the size of a second window until the second local standard deviations are larger than the first local standard deviations, and determining the size of the second window;
the screening unit is used for subtracting the second variogram from the first variogram to obtain a variogram, and selecting a maximum value point in the variogram as a suspected target point;
and the tracking unit is used for tracking the target to be detected according to the suspected target point.
Specifically, the window adjusting unit is configured to adjust a window size, and includes:
the initial sizes of the first window and the second window are N pixel points, and N is more than or equal to 3;
if Sstd_total＞λS1std_localThen the current first window size is determined, otherwise the first window size L = (N + 2N), N =1, 2, 3, … …, Sstd_totalIndicating the global standard deviation, S1std_localDenotes the first local standard deviation, lambda tableShowing an adjusting parameter;
if S2std_local＞λS1std_localThen the current second window size is determined, otherwise the second window size L = (N + 2N), N =1, 2, 3, … …, S2std_localThe second local standard deviation is indicated.
Has the advantages that: compared with the prior art, the invention has the following remarkable advantages: the size change of the target to be detected on the infrared image is determined based on the variance distribution of the obtained image frame of the infrared image, the size of the filtering window is changed in a self-adaptive mode, the weak and small target is stably captured constantly based on the changed filtering window, and the reliability of detecting and tracking the weak and small target is improved.
Drawings
FIG. 1 is a schematic flow chart of an infrared weak and small flying target adaptive detection tracking method provided by the invention;
FIG. 2(a) is a schematic diagram of a first template provided by the present invention; fig. 2(b) is a schematic diagram of a second template provided by the present invention.
Detailed Description
The technical scheme of the invention is further explained by combining the attached drawings.
Referring to fig. 1, a schematic flow chart of an infrared weak and small flying target adaptive detection and tracking method provided by the present invention includes specific steps.
Step 1, acquiring an original infrared image including a target to be detected.
In the specific implementation, the imaging size of the flight target to be measured in the original infrared image generally varies between 1 × 1 to 9 × 9 pixels, and in most cases varies between 4 × 4 to 6 × 6 pixels, although the numerical values themselves are small, the numerical values have a multiple difference. In the prior art, in the actual filtering process, in order to effectively capture the target to be measured, the window size is usually set to be 4 × 4 to 6 × 6 pixels, but when the imaging size of the target changes to 1 × 1 to 2 × 2 or 6 × 6 to 9 × 9, the target is easily lost through the filtering window with the size of 4 × 4 to 6 × 6 pixels.
Step 2, applying a window size adjustment step to the image frame of the original infrared image, calculating by traversing the image frame by using a first template and a second template, and calculating the variance of the gray value in the range of the first template to obtain a first variance map; and calculating the variance of the gray value in the second template range to obtain a second variance map.
In the embodiment of the present invention, the range of the first template is the first window, and the range of the second template is the first window minus the second window located therein.
In specific implementation, a window size adjusting step is applied to an image frame of an original infrared image, the sizes of the filtering windows of the first template and the second template are adjusted, and the adjusted filtering windows are applied to detect an object to be detected for the infrared image.
In a specific implementation, referring to fig. 2(a) and (b), the range of the first template is the whole (hatched portion) of the first window L, and the range of the second template is the portion (hatched portion) of the first window L minus the second window K.
In a specific implementation, the local variance within the template reflects the discrete measure of each gray level distribution in the infrared image, and the local variance distribution highlights detail information of the image. The background of the flying target is sky and cloud layer, the gray level distribution of the target area and the background area is different, so that the variance distribution of the target area and the background area is different, namely the variance of the target area is larger, the variance of the background area is smaller, a part of image is extracted from the background area, the discrete degree of the gray level does not change obviously, and the variance does not change obviously; a part of images are extracted from the target area, the variation of the discrete degree of the gray scale is obvious, and the variation of the variance is also obvious. Therefore, the variance template shown in fig. 2 is established according to the difference between the target area and the background area, and if the second template is within the target area, the variance calculated after excluding the second window will have a more obvious difference from the variance calculated by the first template.
In the embodiment of the invention, the size of the first window and the size of the second window are adjusted by sequentially applying a window size adjusting step to each image frame of the original infrared image.
In a specific implementation, for each image frame, a window size adjustment is performed, the window size adjusted in the previous image frame may be continued to the next image frame, for example, the first window is adjusted to a size of 4 × 4 pixels in the first image frame, and when the window size adjustment is performed on the second image frame, a further adjustment is performed on the basis of the first window having a size of 4 × 4 pixels.
In the embodiment of the invention, the variance of the gray value obtained by calculation is used for replacing the gray value of the center point of the template to obtain the variance graph.
In the embodiment of the present invention, in the second template, the first window and the second window share a common center point.
In the embodiment of the invention, the window size adjusting step comprises the following steps: calculating the global standard deviation of the image frame; traversing the image frame by using a first template, calculating to obtain a plurality of first local standard deviations, if the global standard deviation is not larger than the first local standard deviation, increasing the size of a first window until the global standard deviation is larger than the first local standard deviation, and determining the size of the first window; and traversing the image frame by using a second template, calculating to obtain a plurality of second local standard deviations, if the second local standard deviation is not larger than the first local standard deviation, increasing the size of the second window until the second local standard deviation is larger than the first local standard deviation, and determining the size of the second window.
In particular, the invention provides a scheme for window self-adaption change in order to wrap the target within the window range in real time. Determining the size of a first window according to the size relation between the global standard deviation and the first local standard deviation, wherein when the global standard deviation is larger than the first local standard deviation, the dispersion degree of the gray value in the range of the first window is smaller than the dispersion degree of the gray value in the range of the whole image, which indicates that the first window can accurately capture a specific area; and determining the size of the second window according to the size relation between the first local standard deviation and the second local standard deviation, wherein when the second local standard deviation is larger than the first local standard deviation, the second local standard deviation shows that when a target area is met in the window traversal process and an image with the size of the second window is extracted, the variance obviously changes, and a changed small target can be detected.
In specific implementation, the filtering window is adaptively changed, so that the weak and small targets are stably captured constantly, and the reliability of detecting and tracking the weak and small targets is improved.
In the embodiment of the present invention, the window size adjusting step includes:
the initial sizes of the first window and the second window are N pixel points, and N is more than or equal to 3;
if Sstd_total＞λS1std_localThen the current first window size is determined, otherwise the first window size L = (N + 2N), N =1, 2, 3, … …, Sstd_totalIndicating the global standard deviation, S1std_localDenotes a first local standard deviation, λ denotes a tuning parameter;
if S2std_local＞λS1std_localThen the current second window size is determined, otherwise the second window size L = (N + 2N), N =1, 2, 3, … …, S2std_localThe second local standard deviation is indicated.
In a specific implementation, the initial size of the first window and the second window includes N pixel points, preferably 3 pixel points, and for each image frame, the window size is adjusted, and the window size adjusted in the previous image frame can be continued to the next image frame.
In the specific implementation, the size of the window is adjusted each time, and the 2n pixel points are preferably increased in sequence, so that the size of the window can be gradually increased, and the size of the window is ensured not to be adjusted too large.
In specific implementation, the adjusting parameter λ is added to have a value range of (0, 1), and the relationship between the variances of the first window and the second window can be adjusted.
In the embodiment of the invention, the variance distribution of the gray values of the cloud layer edge area, the cloud layer area and the clearance area is calculated for each image frame of the original infrared image, and the adjusting parameters in the step of adjusting the size of the application window are adjusted according to the variance distribution of each image frame.
In a specific implementation, the background region of the flying target comprises a cloud layer edge region, a cloud layer region and a clearance region, the variance distribution of the gray values in the above specific background region is calculated to indicate the dispersion degree of the background region, if the dispersion degree of the background region is relatively high, the adjustment parameter can be adjusted to be low to overcome the overlarge window caused by the adjustment parameter, and if the dispersion degree of the background region is relatively low, the adjustment parameter can be adjusted to be high to overcome the undersize window caused by the adjustment parameter. The specific discrete degree is relatively higher or relatively lower, based on a preset standard deviation distribution diagram, that is, relatively higher when the degree is higher than the standard deviation distribution diagram, and that is, relatively lower when the degree is lower than the standard deviation distribution diagram.
And 3, subtracting the second variogram from the first variogram to obtain a variogram, and selecting a maximum value point as a suspected target point.
In specific implementation, if the range of the second template is the target area, the variance calculated after excluding the range of the second window will have a significant difference from the variance calculated by the first template, so that the local variance images are subtracted to obtain a local variance difference, and the image represented by the difference represents the target to be measured.
In the embodiment of the invention, the image frame of the original infrared image is subjected to spatial filtering, and the filtered image frame is multiplied by the corresponding pixel point of the variance difference map to obtain a background inhibition map;
screening a preset number of maximum value points in the background inhibition map as suspected target points;
and carrying out target segmentation according to the suspected target point to obtain a suspected target.
In the specific implementation, in order to make the target area more prominent and enhance the contrast ratio between the target and the background area, image multiplication is also one of important improvement points of the invention, and the image after spatial filtering receives effective signals and inhibits interference signals from other directions, and then the effective signals are multiplied by a variance difference chart representing the target, so that the imaging area of the target to be detected can be further highlighted, the target to be detected can be effectively detected and tracked, and the reliability is improved.
In specific implementation, a preset number of maximum value points are selected as suspected target points, wherein the preset number can be set correspondingly according to an actual application scene. Correspondingly, a threshold value can be set, and the maximum value point higher than the threshold value can be used as a suspected target point.
In particular implementations, the target segmentation may take a variety of approaches, such as applying a threshold to segment, and so on.
And 4, tracking the target to be detected according to the suspected target point.
In the embodiment of the invention, a suspected target is tracked by using a wave gate, and if the suspected target is in the wave gate range, the suspected target is determined to be a target to be detected; and if the suspected target is out of the wave gate range, judging the suspected target to be a new target or a false alarm.
The invention also provides an infrared small and weak flying target self-adaptive detection tracking device, which comprises an acquisition unit, a variance calculation unit, a window adjustment unit, a screening unit and a tracking unit, wherein:
the acquisition unit is used for acquiring an original infrared image comprising a target to be detected;
the variance calculating unit is used for applying a window size adjusting step to the image frames of the original infrared images, calculating by traversing the image frames by using a first template and a second template, and calculating the variance of the gray value in the range of the first template to obtain a first variance map; calculating the variance of the gray value in the second template range to obtain a second variance map; the range of the first template is a first window, and the range of the second template is the range of subtracting a second window positioned in the first window from the first window;
the window adjusting unit is used for adjusting the window size: calculating the global standard deviation of the image frame; traversing the image frame by using a first template, calculating to obtain a plurality of first local standard deviations, if the global standard deviation is not larger than the first local standard deviation, increasing the size of a first window until the global standard deviation is larger than the first local standard deviation, and determining the size of the first window; traversing the image frame by using a second template, calculating to obtain a plurality of second local standard deviations, if the second local standard deviations are not larger than the first local standard deviations, increasing the size of a second window until the second local standard deviations are larger than the first local standard deviations, and determining the size of the second window;
the screening unit is used for subtracting the second variogram from the first variogram to obtain a variogram, and selecting a maximum value point in the variogram as a suspected target point;
and the tracking unit is used for tracking the target to be detected according to the suspected target point.
In an embodiment of the present invention, the variance calculating unit is configured to sequentially apply a window size adjusting step to each image frame of the original infrared image to adjust sizes of the first window and the second window.
In the embodiment of the invention, the variance calculating unit is used for replacing the gray value of the central point of the template with the variance of the gray value obtained by calculation to obtain the variance map.
In the embodiment of the invention, the second template, the first window and the second window share a common center point.
In this embodiment of the present invention, the window adjusting unit is configured to adjust a window size, and includes:
the initial sizes of the first window and the second window are N pixel points, and N is more than or equal to 3;
if Sstd_total＞λS1std_localThen the current first window size is determined, otherwise the first window size L = (N + 2N), N =1, 2, 3, … …, Sstd_totalIndicating the global standard deviation, S1std_localDenotes a first local standard deviation, λ denotes a tuning parameter;
if S2std_local＞λS1std_localThen the current second window size is determined, otherwise the second window size L = (N + 2N), N =1, 2, 3, … …, S2std_localThe second local standard deviation is indicated.
In an embodiment of the present invention, the window adjusting unit is configured to calculate a variance distribution of gray values of a cloud layer edge area, a cloud layer area, and a clearance area for each image frame of the original infrared image, and adjust an adjustment parameter in the step of adjusting the size of the application window according to the variance distribution of each image frame.
In the embodiment of the invention, the screening unit is used for performing spatial filtering on the image frame of the original infrared image, and multiplying the filtered image frame by the corresponding pixel points of the variance difference map to obtain a background suppression map;
screening a preset number of maximum value points in the background inhibition map as suspected target points;
and carrying out target segmentation according to the suspected target point to obtain a suspected target.
In the embodiment of the invention, the tracking unit is used for tracking the suspected target by using the wave gate, and if the suspected target is in the wave gate range, the suspected target is determined to be the target to be detected; and if the suspected target is out of the wave gate range, judging the suspected target to be a new target or a false alarm.
Claims (9)
1. An infrared weak and small flying target self-adaptive detection tracking method is characterized by comprising the following steps:
acquiring an original infrared image including a target to be detected;
the image frame of the original infrared image is subjected to window size adjustment, a first template and a second template are suitable for traversing the image frame for calculation, and the variance of the gray value in the range of the first template is calculated to obtain a first variance map; calculating the variance of the gray value in the second template range to obtain a second variance map; the range of the first template is a first window, and the range of the second template is the range of subtracting a second window positioned in the first window from the first window; sequentially applying a window size adjusting step to each image frame of the original infrared image to adjust the sizes of a first window and a second window, wherein the size of the window adjusted in the previous image frame is continued to the next image frame;
wherein, the window size adjusting step: calculating the global standard deviation of the image frame; traversing the image frame by using a first template, calculating to obtain a plurality of first local standard deviations, if the global standard deviation is not larger than the first local standard deviation, increasing the size of a first window until the global standard deviation is larger than the first local standard deviation, and determining the size of the first window; traversing the image frame by using a second template, calculating to obtain a plurality of second local standard deviations, if the second local standard deviations are not larger than the first local standard deviations, increasing the size of a second window until the second local standard deviations are larger than the first local standard deviations, and determining the size of the second window;
subtracting the second variogram from the first variogram to obtain a variogram, and selecting a maximum value point as a suspected target point;
and tracking the target to be detected according to the suspected target point.
2. The method for adaptively detecting and tracking the infrared small and weak flying target according to claim 1, wherein the variance of the gray value in the range of the first template is calculated to obtain a first variance map; calculating the variance of the gray value in the second template range to obtain a second variance map, wherein the second variance map comprises the following steps:
and replacing the gray value of the center point of the template with the calculated variance of the gray value to obtain a variance map.
3. The adaptive infrared small and weak flying target detecting and tracking method as claimed in claim 2, wherein the second template, the first window and the second window are concentric.
4. The infrared weak and small flying target self-adaptive detecting and tracking method according to claim 3, wherein the window size adjusting step comprises:
the initial sizes of the first window and the second window are N pixel points, and N is more than or equal to 3;
if Sstd_total＞λS1std_localThen the current first window size is determined, otherwise the first window size L = (N + 2N), N =1, 2, 3, … …, Sstd_totalIndicating the global standard deviation, S1std_localDenotes a first local standard deviation, λ denotes a tuning parameter;
if S2std_local＞λS1std_localThen the current second window size is determined, otherwise the second window size L = (N + 2N), N =1, 2, 3, … …, S2std_localThe second local standard deviation is indicated.
5. The method as claimed in claim 4, wherein the variance distribution of gray values of the cloud edge region, the cloud region and the clearance region is calculated for each image frame of the original infrared image, and the adjustment parameter in the step of adjusting the size of the application window is adjusted according to the variance distribution of each image frame.
6. The method for adaptively detecting and tracking the infrared small and weak flying target according to claim 5, wherein the selecting the maximum value point as a suspected target point comprises:
carrying out spatial filtering on the image frame of the original infrared image, and multiplying the filtered image frame by corresponding pixel points of the variance difference image to obtain a background inhibition image;
screening a preset number of maximum value points in the background inhibition map as suspected target points;
and carrying out target segmentation according to the suspected target point to obtain a suspected target.
7. The method for adaptively detecting and tracking the infrared small and weak flying target according to claim 6, wherein the tracking the target to be detected according to the suspected target point comprises:
tracking the suspected target by using a wave gate, and if the suspected target is in the wave gate range, determining the suspected target as a target to be detected; and if the suspected target is out of the wave gate range, judging the suspected target to be a new target or a false alarm.
8. The utility model provides an infrared weak little flying object self-adaptation detects tracking means which characterized in that, includes acquisition element, variance calculating unit, window adjustment unit, screening unit and tracking unit, wherein:
the acquisition unit is used for acquiring an original infrared image comprising a target to be detected;
the variance calculating unit is used for applying a window size adjusting step to the image frames of the original infrared images, calculating by traversing the image frames by using a first template and a second template, and calculating the variance of the gray value in the range of the first template to obtain a first variance map; calculating the variance of the gray value in the second template range to obtain a second variance map; the range of the first template is a first window, and the range of the second template is the range of subtracting a second window positioned in the first window from the first window; sequentially applying a window size adjusting step to each image frame of the original infrared image to adjust the sizes of a first window and a second window, wherein the size of the window adjusted in the previous image frame is continued to the next image frame;
the window adjusting unit is used for adjusting the window size: calculating the global standard deviation of the image frame; traversing the image frame by using a first template, calculating to obtain a plurality of first local standard deviations, if the global standard deviation is not larger than the first local standard deviation, increasing the size of a first window until the global standard deviation is larger than the first local standard deviation, and determining the size of the first window; traversing the image frame by using a second template, calculating to obtain a plurality of second local standard deviations, if the second local standard deviations are not larger than the first local standard deviations, increasing the size of a second window until the second local standard deviations are larger than the first local standard deviations, and determining the size of the second window;
the screening unit is used for subtracting the second variogram from the first variogram to obtain a variogram, and selecting a maximum value point in the variogram as a suspected target point;
and the tracking unit is used for tracking the target to be detected according to the suspected target point.
9. The adaptive detection and tracking device for the infrared small and weak flying target according to claim 8, wherein the window adjusting unit is used for adjusting the size of the window, and comprises:
the initial sizes of the first window and the second window are N pixel points, and N is more than or equal to 3;
if Sstd_total＞λS1std_localThen the current first window size is determined, otherwise the first window size L = (N + 2N), N =1, 2, 3, … …, Sstd_totalIndicating the global standard deviation, S1std_localDenotes a first local standard deviation, λ denotes a tuning parameter;
if S2std_local＞λS1std_localThen determineA current second window size, otherwise the second window size L = (N + 2N), N =1, 2, 3, … …, S2std_localThe second local standard deviation is indicated.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111142157.4A CN113592914B (en) | 2021-09-28 | 2021-09-28 | Infrared weak and small flying target self-adaptive detection tracking method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111142157.4A CN113592914B (en) | 2021-09-28 | 2021-09-28 | Infrared weak and small flying target self-adaptive detection tracking method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113592914A CN113592914A (en) | 2021-11-02 |
| CN113592914B true CN113592914B (en) | 2022-04-01 |
Family
ID=78242368
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111142157.4A Active CN113592914B (en) | 2021-09-28 | 2021-09-28 | Infrared weak and small flying target self-adaptive detection tracking method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113592914B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117218161B (en) * | 2023-11-09 | 2024-01-16 | 聊城市敏锐信息科技有限公司 | Fish track tracking method and system in fish tank |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109767439A (en) * | 2019-01-10 | 2019-05-17 | 中国科学院上海技术物理研究所 | A kind of multiple dimensioned difference of self-adapting window and the object detection method of bilateral filtering |
| CN111353496A (en) * | 2018-12-20 | 2020-06-30 | 中国科学院沈阳自动化研究所 | Real-time detection method for infrared small and weak target |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP3140781B1 (en) * | 2014-05-08 | 2019-06-12 | Aselsan Elektronik Sanayi ve Ticaret Anonim Sirketi | A real- time, semi-automatic method for target tracking window initialization in thermal imagery |
-
2021
- 2021-09-28 CN CN202111142157.4A patent/CN113592914B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111353496A (en) * | 2018-12-20 | 2020-06-30 | 中国科学院沈阳自动化研究所 | Real-time detection method for infrared small and weak target |
| CN109767439A (en) * | 2019-01-10 | 2019-05-17 | 中国科学院上海技术物理研究所 | A kind of multiple dimensioned difference of self-adapting window and the object detection method of bilateral filtering |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113592914A (en) | 2021-11-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109633633B (en) | Life signal enhancement method based on segmented classification enhancement processing | |
| JP3123587B2 (en) | Moving object region extraction method using background subtraction | |
| CN111476809A (en) | Side-scan sonar image target identification method | |
| CN109740445B (en) | Method for detecting infrared dim target with variable size | |
| CN110728697A (en) | Infrared dim target detection tracking method based on convolutional neural network | |
| CN109584204A (en) | A kind of image noise intensity estimation method, storage medium, processing and identification device | |
| CA2780595A1 (en) | Method and multi-scale attention system for spatiotemporal change determination and object detection | |
| EP2846306B1 (en) | Image processing apparatus for removing haze contained in still image and method thereof | |
| CN107274410B (en) | Self-adaptive artificial target constant false alarm rate detection method | |
| CN113592914B (en) | Infrared weak and small flying target self-adaptive detection tracking method and device | |
| RU2419150C1 (en) | Method to process sequence of images to detect and follow air objects | |
| CA2704037A1 (en) | Method for detecting a target | |
| KR101956250B1 (en) | Coastline monitoring apparatus and method using ocean color image | |
| CN111612773A (en) | Thermal infrared imager and real-time automatic blind pixel detection processing method | |
| CN107610056B (en) | Mixed weighting wiener filtering image denoising method based on total variation | |
| CN114782561B (en) | Smart agriculture cloud platform monitoring system based on big data | |
| CN107563981B (en) | Method for evaluating image ringing effect | |
| Kumar et al. | Image enhancement and performance evaluation using various filters for IRS-P6 Satellite Liss IV remotely sensed data | |
| Lu et al. | Research on rainfall identification based on the echo differential value from X-band navigation radar image | |
| CN114429593A (en) | Infrared small target detection method based on rapid guided filtering and application thereof | |
| CN113705672A (en) | Threshold value selection method, system and device for image target detection and storage medium | |
| JP6172641B2 (en) | Image filtering apparatus, image filtering method, and image filtering program | |
| CN112233022A (en) | Laser radar guided filtering denoising method based on gradient domain correction | |
| Tiwari et al. | Estimation based non-local approach for pre-processing of MRI | |
| CN117710245B (en) | Astronomical telescope error rapid detection method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |